Nvidia’s reported move to halve the number of Asian customers authorised to buy its AI chips is more than a compliance story. It is a blunt reminder that Southeast Asia’s position in the global AI supply chain is neither neutral nor secure. The region is increasingly being treated not as a frontier for innovation alone, but as a possible circuit board in a much larger geopolitical struggle.

According to the Financial Times, Nvidia has tightened due diligence across Singapore, Malaysia, and Japan, removing more than half of its previous customers from an internal white list after tougher checks failed to clear many of them.

Also Read: Why Asia sits at the centre of the global AI chip disruption?

The obvious explanation is export control pressure from Washington, especially as the US tries to stop advanced chips from leaking to China through third countries. The less comfortable truth is that Southeast Asia has become a test case for how much trust global technology giants are willing to extend to local buyers.

This is not just about chips; it is about trust

For years, the region’s tech ecosystem has benefited from a simple assumption: if you can pay, you can play. That era is ending. In its place comes a more suspicious age in which companies are not merely customers, but potential compliance risks to be vetted, interviewed and visited in person. Data centres are inspected, contracts are checked, end users are questioned. The logic is not glamorous, but it is powerful.

For Southeast Asian companies, especially the neo-cloud providers that depend on Nvidia hardware to sell AI infrastructure, this is a serious blow. These businesses have marketed themselves as agile alternatives to the hyperscalers, offering access to scarce compute capacity in markets hungry for AI experimentation. Many have thrived on the promise that the region could become a genuine hub for distributed AI infrastructure, not just a consumer of imported technology.

Now they are being asked a more awkward question: are you a legitimate AI business, or a convenient waypoint in a sanctions-bypassing supply chain?

That question is important because, in Southeast Asia, perception can quickly become policy. Once a market is associated with trans-shipment concerns, the bar for participation rises sharply. Legitimate firms get dragged into the same scrutiny as bad actors. The result is a kind of collateral distrust. It is not an outright ban, but it can feel like one if you are the company suddenly trying to explain why your servers, customers and contracts look exactly as opaque as everyone feared they might.

Singapore and Malaysia are in the spotlight for different reasons

Singapore’s tech sector will feel this differently from Malaysia’s, but neither gets to escape the consequences. Singapore has spent years positioning itself as the region’s clean, well-regulated digital hub: the place where serious cloud players, AI labs and semiconductor investors can do business with confidence. If Nvidia’s checks are now focusing heavily on compliance in Singapore, that is not a compliment. It is a sign that even the most institutionally trusted markets are being pulled deeper into the enforcement perimeter.

Malaysia, meanwhile, sits closer to the hard edge of the issue. Its data centre boom has been one of the region’s most exciting investment narratives, with land, power and regional connectivity attracting a wave of global attention. But any boom built on the assumption of frictionless access to leading-edge chips is vulnerable when geopolitics decides to become a gatekeeper.

Also Read: Asia rises in the AI chip race: China to outgrow US by 30 per cent by 2030

The irony is hard to miss. Southeast Asia is simultaneously being asked to build more digital infrastructure and to prove that this infrastructure will not be used in ways Washington dislikes. That is a tall order for a region that has historically preferred strategic ambiguity. Ambiguity is useful for diplomacy. It is less useful when your supplier wants names, use cases, contracts and the moral character references of your end users.

The AI race is becoming a compliance race

There is another uncomfortable lesson here: in AI, access to compute is now as strategic as access to capital. Nvidia’s chips are not just components; they are the toll gates of the modern AI economy. Whoever controls access controls the pace of development. And when those gates narrow, the impact is uneven.

Large enterprises and hyperscalers may absorb the shock. Smaller companies cannot. Startups building AI products, niche cloud providers and regional infrastructure players often depend on predictable supply and fast procurement. A white list turns supply from a business decision into a political and procedural one. That slows expansion, raises costs and makes planning harder. In short, it turns growth into a paperwork sport.

This matters because Southeast Asia is still trying to prove that it can produce AI companies, not merely host AI servers. If access to top-tier chips becomes more selective, the region’s emerging players may find themselves competing not just on product quality, but on the sophistication of their compliance teams. The irony is exquisite, and somewhat depressing: the future of AI might depend on who can produce the most convincing audit trail.

Washington’s shadow is widening

The deeper issue is that US policy is no longer simply about banning exports to China. It is about shaping the behaviour of third countries and private companies far beyond America’s borders. That is what makes this move so consequential for Southeast Asia. The region is not the target, but it is increasingly part of the mechanism.

The Commerce Department’s May guidance, aimed at advanced AI chips reaching overseas subsidiaries of Chinese companies, signals a broader enforcement mindset: if there is a route around the wall, the wall will be extended. Nvidia’s reported inspections and end-user interviews are the corporate translation of that policy logic. The company is not acting in a vacuum; it is trying to stay ahead of the regulator by turning compliance into a product feature.

This leaves Southeast Asian firms in a difficult position. They are expected to behave like sophisticated global operators, but many are still maturing operationally. Some may indeed have weak controls or murky customer links. Others may simply lack the legal, governance and documentation infrastructure demanded by American vendors in an era of intense scrutiny. Either way, the burden falls on the local ecosystem to prove innocence in advance.

What happens next will shape the region’s AI market

The immediate market effect will likely be consolidation. Firms that can clear compliance hurdles will gain advantage; those that cannot may lose access to Nvidia hardware or face delays that wreck business plans. Some will rebrand, restructure or cut ties with questionable clients. Others will disappear into the long list of regional companies that once looked promising until geopolitics discovered them.

But there is also a longer-term possibility: this shock could force Southeast Asia to professionalise faster. Better governance, cleaner customer due diligence and clearer ownership structures are not glamorous, but they are the price of admission to the high-end AI economy. The region cannot build a serious AI industry on hand-waving and optimism alone. The chip wars have ended that fantasy.

Also Read: Building smart: A tech founder’s guide to the semiconductor supply chain revolution

Still, there is a risk of overcorrection. If compliance becomes so heavy-handed that only the largest and safest buyers can participate, the region could end up with a concentrated AI market that serves incumbents and excludes the very startups most likely to drive innovation. That would be a classic Southeast Asian tragedy: enormous potential, strangled by asymmetric rules written elsewhere.

Nvidia’s white list may be a technical adjustment, but it carries a strategic message. Southeast Asia is no longer operating in a benign global market. It is operating in a filtered one. And in this new world, access to AI compute is not just a commercial advantage. It is a politically contested privilege.

The post The Nvidia clampdown is a warning for Southeast Asia’s AI boom appeared first on e27.

As AI commoditises everything a company makes, the last defensible moat is not what you sell — it is the human experience you design around it. Most companies are investing in exactly the wrong thing.

In the mid-1990s, a Nike marketer told a room of executives: “We don’t sell shoes. We sell the feeling of being an athlete.” Three decades on, it reads like strategy. Walk into almost any product review today — specifications, roadmaps, feature releases. What the company makes. Rarely does the customer become.

Yesterday, a marketer’s Instagram Reel stopped my scroll. Twenty-nine likes. Just this: “People don’t pay for skincare. They pay to feel confident walking into a room. They don’t pay for coaching. They pay for certainty of achieving a goal.” You knew this already. So why does your board deck open with product metrics — not with who your customer is trying to become?

The milkshake nobody understood

In the late 1990s, a fast-food chain hired consultants to fix flat milkshake sales. Surveys. Focus groups. Flavour tests. Nothing moved. Then a researcher did something different: he watched. The most reliable customer was a lone commuter before 8 am, long drive ahead — not buying sweetness, but hiring something to defeat boredom. A banana was gone in two bites; a doughnut left sticky fingers on the wheel. The milkshake lasted twenty minutes. The competitor was not Burger King. It was the commute itself. The chain had spent months asking the wrong question.

What that researcher practised was radical empathy — not asking customers what they wanted, but inhabiting their experience long enough to see what they could not say. Clayton Christensen built Jobs to Be Done around this. His arithmetic was unsparing: 75 to 85 per cent of new products fail — not from poor execution, but because companies never understood the job the customer needed done. A concurrent McKinsey survey found nine in ten global executives dissatisfied with their innovation results. Nine in ten — after all the data, all the research, all the frameworks. The data exists. The empathy does not.

“Frame your business around the products you sell, and you get supplanted when technology changes. Frame it around the job you do, and new technologies become tools to do it better.”— Clayton Christensen, Competing Against Luck, 2016

What the East knew first

You might think this is what CRM systems are for. What recommendation engines do. What personalisation at scale delivers. The Japanese figured this out centuries before the algorithm — and arrived at something entirely different.

The word is omotenashi. Western management translates it as “hospitality.” That is not right. Hospitality responds. Omotenashi anticipates. Service gives you what you ask for. Omotenashi ensures you never have to ask. At Isetan in Tokyo, umbrella bags appear at the entrance before you notice you need one. No complaint triggered this. No model predicted it. Someone simply asked: What will this person feel when they walk in? That question — not the algorithm — is human experience design.

Also Read: When AI leaves the screen, cybersecurity becomes product responsibility

One framework came from Harvard. The other predates the printing press. Same conclusion — which most boardrooms still treat as optional: the organisation that wins understands what the customer has not yet found the words to say. Your CRM cannot do that. Only radical empathy can.

When the story collapses

Peloton is the case study nobody wants to be. In December 2020, its stock touched US$162 — a US$2,500 bicycle turned into a cultural identity: not hardware, but the sensation of being a serious athlete, accountable to a tribe. By January 2022: US$24. Most analysts blamed the reopened gyms. Wrong. The product had not changed. The instructors had not left. What collapsed was the story customers told about themselves when they used it. Peloton had never designed that story — they had stumbled into it. When the context shifted, there was nothing to hold it in place.

Apple made the opposite bet, deliberately. Jobs redesigned the Apple Store around one question: not what do people come here to buy, but what do they come here to become? The result was human experience design in its purest form — not a product environment, but an encounter with a more capable self. That encounter cannot be copied or shipped in a software update. It lives in the designed space between a brand and a human being — which is, not coincidentally, why Apple’s retail revenue per square foot still leads every category.

The speed at which AI commoditises what companies make will always outpace the speed at which companies learn to understand what people feel. Radical empathy is not a corrective. It is the only strategy left.

The trap of intelligent personalisation

Here is what most AI transformation roadmaps assume: that personalisation at scale is omotenashi. It is not. Omotenashi is radical empathy — unhurried observation of one specific person in one unrepeatable moment. AI personalisation is pattern-matching: the customer receives what people like them statistically want. That is not empathy. That is a fast guess with good data. Customers can feel the difference between being understood and being predicted.

Zurich Insurance ran the experiment. Between 2023 and 2025, more than a quarter of its workforce completed empathy training — 46,000 hours. Its Net Promoter Score rose seven points in eighteen months: not from a product launch or a price cut, but from understanding what a customer was actually feeling. The ROI of radical empathy was not soft. It was the only lever that moved.

Also Read: Seasonal product cycles: Why some features only work at certain times

Accenture’s 2025 Life Trends study found consumers in 22 markets accumulating a “cost of hesitation” — rising distrust of digital content, hunger for something real. When everything can be generated, authenticity becomes the scarce good. Your competitors have the same models. They are running the same optimisations. What they cannot replicate is the human experience you choose to design.

The thing you have not built

Starbucks did not lose a decade because the coffee got worse. It lost the third place — that feeling that the room belonged to you — the moment efficiency became the priority. The product survived. The experience did not. Most leadership teams, hearing this, nod. Then return to optimising throughput.

The most defensible asset a company can build is not a product. It is the story customers tell about themselves when they choose you. That story cannot be generated. It cannot be A/B tested into existence. It has to be designed — through radical empathy, one human experience at a time. Most organisations have more customer data than at any point in history. They understand their customers less than they did a decade ago. That is not a paradox. It is what happens when measurement becomes the goal, and the thing being measured gets forgotten.

Your company has a Chief Data Officer. Probably a Chief AI Officer. Perhaps a Chief Experience Officer. When did any of them last spend an unscripted hour inside a customer’s actual day — not an interview, not a dashboard, just watching what their life costs them? If that question requires thought, you already know what is missing.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.

The post Your customers are not buying your product, they are buying a better version of themselves appeared first on e27.

The United States digital asset ecosystem faces a precarious turning point today as sweeping regulatory changes collide with severe macroeconomic shocks. The Securities and Exchange Commission recently advanced its extensive Regulation Crypto rule package to formal White House review. This comprehensive market framework moves the federal government away from case-by-case enforcement actions toward a predictable regime for token distribution and capital formation. This regulatory progression arrives exactly as intense macroeconomic headwinds reshape digital asset valuations in real time.

The entire cryptocurrency market capitalisation contracted by 2.07 per cent to US$2.15T in a single 24-hour period. This rapid decline highlights the extreme vulnerability of speculative assets when geopolitical instability strikes the global financial system. During this period of uncertainty, digital assets showed an 87 per cent correlation with traditional gold. This strong correlation indicates that investors are increasingly treating top-tier cryptocurrencies as traditional macro hedges during periods of sudden international tension.

The movement of the 400+page Regulation Crypto document into the Office of Information and Regulatory Affairs initiates an imminent pre-publication phase. Once the executive branch concludes this comprehensive scrutiny, the regulatory agency will publish the extensive text in the Federal Register. This publication will trigger a formal public comment period where industry participants and lawmakers can actively lobby for critical modifications.

The commission designed this regulatory push to work alongside the CLARITY Act currently before Congress. While the pending congressional legislation explicitly divides market oversight between the commodity and securities watchdogs, the administrative rulemaking focuses primarily on practical fundraising pathways.

If the legislative path remains gridlocked, the federal administrative agency will likely establish this rulebook as the default framework for domestic digital assets. Prominent political figures, particularly Senate Democrats, already argue that the commission is attempting to legislate through administrative rules rather than waiting for congressional authorisation.

Also Read: Why Bitcoin’s move to US$63K has nothing to do with crypto and everything to do with Iran

The capital formation proposals embedded within the new framework have the potential to radically alter how digital asset projects raise capital within the United States. The proposed text outlines a four-year startup exemption that permits emerging projects to raise up to US$5,000,000 annually using structured disclosures rather than full registration.

For mature issuers, a secondary tier allows capital raises of up to US$75,000,000 per year under a significantly lighter regulatory burden. The framework also introduces a vital investment contract safe harbour. This safe harbour provides a clear mechanism for a token to exit its classification as a security once the original issuer permanently concludes all manager-led efforts.

The agency builds this concept on a joint taxonomy that distinguishes among digital commodities, collectibles, tools, stablecoins, and securities. This taxonomy establishes a baseline presumption that most tokens do not qualify as securities unless issuers explicitly market and sell them as investment contracts. These rules give domestic projects much clearer paths to raise capital and could reopen domestic funding channels that previously moved offshore.

Despite these structural regulatory developments, the immediate valuation of digital assets remains highly vulnerable to broader macroeconomic shocks and sudden liquidation cascades. A geopolitical risk-off cascade rippled across global asset classes and primarily drove the recent market decline.

Reports of United States and Iran military strikes over the Strait of Hormuz on 13 July caused global equities to retreat sharply. This macro-driven contraction rapidly translated into a violent unwind of speculative leverage across cryptocurrency spot and derivatives markets. The sudden panic forced over US$95,870,000 in Bitcoin liquidations within 24 hours. Long positions accounted for a staggering 92 per cent of that wiped-out capital.

The market absorbed a devastating one-two punch as the macroeconomic shock eroded risk appetite and forced selling from overleveraged positions, aggressively accelerating the downward price action. Traders must now watch for any de-escalation in geopolitical headlines because easing tensions could quickly relieve the immense selling pressure.

Also Read: The Independence Day crypto puzzle: Up or down?

The short-term trajectory for digital assets remains highly sensitive to incoming economic data and subsequent monetary policy responses. Market participants have adopted a deeply cautious stance ahead of the 14 July United States Consumer Price Index report. Traders fear that a hot inflation print will force the Federal Reserve to maintain its restrictive monetary policy and hawkish rhetoric. The digital asset market capitalisation is currently testing a critical Fibonacci support level at US$2.13T, which represents a 61.8 per cent retracement.

A benign inflation report could stabilise prices and spark a steady rebound toward the recent US$2.15T pivot. An uncomfortably high inflation reading risks accelerating the sell-off toward a direct retest of the yearly low at US$2.04T. The persistent regulatory overhang from recent agency classifications compounds this negative sentiment. The complete absence of immediate positive catalysts leaves the valuation landscape highly exposed to these incoming external economic indicators.

The underlying price action of Bitcoin highlights the ongoing conflict between retail panic and deep-pocketed buyers. Bitcoin experienced a significant downward flush that terrified average investors before staging a rapid recovery to climb back above the US$62,000 threshold. The price range between US$60,000 and US$61,000 has been a major technical battleground for several months. The brief drop below this accumulation zone acted as a classic bear trap.

The last time the asset dropped below this critical level, the market crashed to the US$58,000 mark. By reclaiming the US$62,000 level, buyers successfully absorbed the immediate selling pressure and removed the initial selling force from the market. The daily time frame’s structural outlook looks significantly healthier, with Bitcoin trading above this zone. The next major horizontal overhead resistance sits at approximately US$64,000. Clearing that specific hurdle will likely spark a quick return to all-time highs because sellers appear to be running out of momentum.

Also Read: Why the 4.1% PCE inflation print just turned crypto into a high beta risk asset

Underlying network metrics further validate the resilience of the primary cryptocurrency and reveal that smart money is actively accumulating during this market weakness. While regulated spot exchange-traded funds have experienced visible bleeding, on-chain tracking metrics paint a completely different picture behind the curtain. Whale addresses completely ignored the retail panic and refused to sell during the sudden price drop.

A recent Bitfinex report indicates that addresses holding 1,000 BTC or more aggressively expanded their holdings over a brief two-week window. These massive entities added more than 270,000 BTC to their vaults while the spot premium remained highly volatile. This massive accumulation represents an influx of more than US$16,700,000,000 in purchasing power. This data clearly demonstrates that sophisticated entities are treating the geopolitical panic and regulatory uncertainty as a generational buying opportunity.

The convergence of massive institutional accumulation, macroeconomic volatility, and shifting regulatory boundaries suggests that the domestic digital asset industry is entering a highly mature phase. The era of regulation by enforcement is gradually yielding to a structured regime that favours capitalised issuers capable of navigating complex legal systems. The impending legal battles will determine whether the executive branch can successfully implement these sweeping changes and provide the clarity that institutional capital demands.

Retail investors continue to obsess over daily fluctuations driven by international conflict and inflation prints. The largest entities in the space are quietly establishing massive architectural positions in anticipation of a fully regulated future. This ongoing transfer of wealth from panicked retail traders to convicted institutional holders will ultimately dictate the next major expansion cycle for the entire digital asset ecosystem.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.

The post Regulation crypto is here: The 400-page rule that could kill or save American crypto innovation appeared first on e27.

PixVerse has extended its Series C round, bringing total fundraising in the round to US$439 million, as the AI video generation company seeks to move beyond short-form video creation into games, virtual hosting, and real-time interactive entertainment.

The company did not disclose the size of the extension, its latest valuation, or how the round was structured. New investors in the extension include Alibaba, Lollapalooza Capital, Ivy Capital, Grand Mount Capital, Eastern Bell Capital, Mirae Asset, BlueFocus, and CloudAlpha.

Also Read: PixVerse’s Jaden X: Why AI video’s biggest opportunity isn’t Hollywood

Existing backers iGlobe Partners and OCBC’s LionX Ventures also participated.

Founded in 2023, PixVerse says it has more than 150 million users across over 177 countries. Its core product allows users to generate video from text prompts, photos or clips, a market that has become crowded as OpenAI, Google, Runway, Luma, Pika, Kuaishou’s Kling and MiniMax compete to define the next interface for digital content creation.

The new funding comes as AI video companies face a narrowing window to show that they can become platforms rather than features. Text-to-video models have improved quickly, but distribution, cost, copyright risk and monetisation remain open questions. PixVerse’s answer is to push into interactive worlds, where users do not just generate a clip but shape an environment in real time.

From generated clips to generated worlds

At the centre of PixVerse’s expansion is R1, which the company describes as a real-time world model. Launched in January 2026, R1 is designed to produce a continuous interactive video stream that responds to user input rather than a fixed rendered output.

PixVerse later added shared worlds and personalised avatars, allowing multiple users to enter the same AI-generated environment. The company now plans to apply that capability to game creation and livestream entertainment.

“The world a player inhabits is not pre-rendered but continuously generated, in real time, in response to what they do,” said Changhu Wang, co-founder and CEO of PixVerse. “That is a fundamentally different foundation for what a game can be.”

That is the right strategic direction, but also the more difficult one. Generating a compelling video clip is one problem; generating a playable, coherent and persistent world is another. Games require rules, balance, latency control, memory, moderation, asset consistency and user retention. The technical bar is higher, and so is the commercial one.

PixVerse says its Game Engine separates game mechanics from visual expression. In practice, this means a creator can define rules or structures in natural language, while the system generates the world and its visual responses in real time. If it works at scale, this could lower the barrier for non-technical creators who want to build lightweight game experiences without modelling, animation or scripting expertise.

Why Southeast Asia matters

The Southeast Asian relevance is clear. The region is one of the world’s most active mobile-first entertainment markets, with a young population, high social media usage and strong creator adoption. Google, Temasek and Bain estimated Southeast Asia’s digital economy gross merchandise value at about US$263 billion in 2024, while Niko Partners has estimated that Southeast Asia and Taiwan together had more than 330 million gamers and a games market worth several billion US dollars.

Also Read: It is time to democratise video-making. Here is how we are going to help the cause

The region is also structurally suited to experiments in AI-generated entertainment. Indonesia, Vietnam, Thailand, and the Philippines have large creator communities and deep mobile gaming cultures. Singapore provides a hub for capital, AI governance and regional headquarters. Malaysia and Vietnam have growing game development and outsourcing talent pools. For platforms such as PixVerse, Southeast Asia is not merely a user acquisition market; it is a testing ground for social, mobile and creator-led use cases.

There is also a live-streaming angle. E-commerce and entertainment live-streaming have become mainstream across markets, such as Indonesia, Thailand and Vietnam through TikTok Shop, Shopee Live and LazLive. AI-generated characters that can respond to viewers in real time may find early commercial use in virtual hosting, fan engagement and branded entertainment. That said, regional regulators are increasingly sensitive to synthetic media, advertising disclosures and platform accountability.

Competitive pressure is rising

PixVerse’s move into games places it closer to a different set of competitors. In gaming infrastructure, Unity and Epic Games’s Unreal Engine remain dominant. In AI-assisted game creation, companies such as Inworld AI, Scenario, Rosebud AI and several asset-generation startups are trying to automate character behaviour, world-building and production workflows.

In Southeast Asia, the competitive context is more fragmented. Sea Group’s Garena, Vietnam’s VNGGames, Indonesia’s Agate, and Singapore-based studios such as Mighty Bear Games have built regional experience in mobile and online games. These companies may not compete directly with PixVerse at the model layer, but they understand what regional players actually spend time and money on: social loops, multiplayer mechanics, localisation and distribution.

PixVerse will also need to prove economics. Real-time generation can be expensive, particularly if users expect low latency and high visual quality. The company’s large user base gives it distribution, but usage does not automatically translate into durable revenue. AI video platforms have often benefited from viral experimentation; games demand repeat engagement.

The investor list suggests PixVerse is positioning itself across several strategic lanes. Alibaba brings distribution and cloud relevance. BlueFocus has advertising and marketing services exposure. Mirae Asset and OCBC-linked LionX Ventures add financial and regional institutional weight. The presence of iGlobe Partners also reinforces the Singapore and cross-border venture connection.

The next test

PixVerse’s funding extension underscores how quickly generative AI companies are being pushed to expand their ambitions. The first wave focused on replacing or compressing creative production workflows. The next wave is trying to create new formats that were previously impractical: persistent AI worlds, adaptive characters, and interactive media that respond to audiences in real time.

For Southeast Asia, that shift could matter if it gives creators and small studios cheaper tools to build interactive content for mobile-first audiences. It could also intensify concerns around copyright, labour displacement, moderation and synthetic identity, particularly in markets where platform regulation is still catching up.

Also Read: Generative AI: The unstoppable force reshaping work and engagement across SEA

PixVerse has raised enough capital to compete seriously. The harder question is whether it can turn real-time world generation from a technical claim into a product that creators, players and entertainment businesses use repeatedly. In AI video, novelty travels fast. In games, only retention counts.

The post Alibaba, Mirae Asset join PixVerse’s US$439M Series C push into AI games appeared first on e27.

Southeast Asia’s startup ecosystem has spent the past several years wrestling with a familiar problem: founders and enterprises alike struggle to find the right partners at the right time. Corporates looking to work with startups often lack visibility into who is building what, while startups seeking distribution, capital, or co-development partners frequently rely on informal networks, chance encounters, or the same recycled circle of contacts. As open innovation becomes a strategic imperative rather than a nice-to-have, the region needs infrastructure that can systematically connect the right organisations to each other, at scale and with intent.

This is the gap e27 set out to address when it introduced its AI-powered business matchmaking platform at Echelon Singapore 2026, a tool designed to help founders, investors, and enterprises identify and engage with relevant counterparts across the region’s fast-moving ecosystem. The initial rollout demonstrated strong appetite among event participants for a more structured, data-driven approach to partnership discovery, one that goes beyond business cards and hallway conversations.

Building on that momentum, e27 is now expanding the reach and capability of its matchmaking platform through a new collaboration with Sony Acceleration Platform, the entity behind Boundary Spanning Service, a business matching platform that has already built meaningful traction in Japan since its launch in January 2025. The collaboration brings together two ecosystems, e27‘s Southeast Asian startup and enterprise network, starting with Singapore and Sony Acceleration Platform’s roster of Japanese corporates and ventures, under a shared goal of making cross-border business matching more efficient and outcome-driven.

Matchmaking built for how founders actually work

e27‘s AI-powered Business Matchmaking platform was built around a simple premise: partnership discovery should not depend on luck. By using AI to parse the profiles, needs, and strengths of registered organisations, the platform surfaces relevant connections that founders and enterprises might otherwise never encounter.

This matters in a region as fragmented and fast-growing as Southeast Asia, where startups and corporates are often only a few relationships away from a breakthrough partnership, joint venture, or sales channel, yet lack an efficient way to identify who those relationships should be.

Also Read: Report: Asia Pacific, Japan drive the next wave of global AI innovation

The platform’s debut at Echelon Singapore 2026 gave founders, operators, and enterprise representatives a first look at how AI can support the earliest and often most difficult stage of any collaboration: figuring out who to talk to. For a region where digital transformation, deep tech, and enterprise innovation are increasingly cross-border pursuits, that kind of structured discovery layer is becoming essential infrastructure rather than a convenience.

Extending the network into Japan through Sony Acceleration Platform

The collaboration with Sony Acceleration Platform gives this vision considerably more depth. Boundary Spanning Service, the matchmaking platform Sony Acceleration Platform operates, has registered around 1,900 organisations and several thousand users based in Japan as of June 2026, spanning both large corporates and startups.

On the enterprise side, its network includes general trading companies, banks and insurers, electronics and precision instrument makers, real estate, chemistry, telecommunications, food, and transportation equipment players, alongside numerous divisions within Sony Group itself, from Sony Semiconductor Solutions and Sony Interactive Entertainment to Sony Music and Sony Financial Group. On the startup side, registered ventures span AI, healthcare, biology, deep tech, VR, data analytics, digital, and SNS marketing.

By connecting e27‘s Southeast Asian matchmaking infrastructure with Sony Acceleration Platform’s Japan-based network, the collaboration is designed to widen the aperture for founders and enterprises on both sides. Starting with Singapore, startups gain a more direct path to engaging Japanese corporates and ventures for joint research and development, sales expansion, or new business exploration, while Japanese organisations gain visibility into a Southeast Asian startup landscape that is increasingly relevant to their own innovation and growth strategies.

Early user feedback shared by Sony Acceleration Platform, including from companies that have used Boundary Spanning Service to secure meetings and identify co-creation partners shortly after registering, points to the kind of tangible engagement this expanded network aims to replicate at a cross-border level.

Also Read: Inside the AI Workflow Competition at Echelon Singapore 2026

Why this collaboration matters now

For founders and enterprise teams navigating an increasingly interconnected but still fragmented regional and cross-border ecosystem, this collaboration is a signal of where business matchmaking is headed: less reliant on manual networking, more structured around data and shared platforms that span multiple markets.

Organisations engaging with e27 and Sony Acceleration Platform through this expanded network can expect a more systematic route to identifying partners for joint development, distribution, or investment conversations, whether they are a Singapore-based startup eyeing the Japanese market or a Japanese enterprise scouting innovation from the region.

As AI-driven tools become more embedded in how partnerships are sourced and vetted, collaborations like this one point to a broader shift in how ecosystems connect across borders. Matchmaking is moving from an occasional, event-driven activity to an ongoing, technology-supported process, and platforms that can bridge distinct regional networks stand to play an outsized role in shaping which collaborations actually get off the ground.

Southeast Asia’s startup and technology ecosystem continues to mature quickly, and cross-border collaboration is becoming less of an exception and more of an expectation. The expansion of e27‘s AI-powered matchmaking platform through this collaboration with Sony Acceleration Platform reflects that trajectory, giving founders and enterprises on both sides a clearer, more direct route to the collaborations that matter.

The region is evolving quickly, and e27 offers the right place at the right moment to be part of what comes next.

Register here to join the conversation.

—

The post e27 expands AI-powered business matchmaking with Sony Acceleration Platform collaboration appeared first on e27.

Ransomware groups are no longer hunting for the biggest fish. They are hunting for the most defenceless ones, and across Southeast Asia, that means small and medium enterprises (SMEs).

The shift is deliberate. As large corporations harden their defences with enterprise-grade security stacks, threat actors have recalibrated their playbooks, zeroing in on SMEs that often run outdated software, lack dedicated IT security staff, and treat cybersecurity as a cost centre rather than a business imperative.

The extortion tactics have also grown more sophisticated: double and triple extortion — where attackers encrypt data, threaten public exposure, and target a firm’s clients or partners — are now standard operating procedure.

For Southeast Asia, the exposure is acute. The region’s SME backbone powers a significant share of GDP across markets like Indonesia, Vietnam, and the Philippines, yet digital literacy and cyber readiness remain uneven. Many businesses digitalised rapidly during the pandemic without the security foundations to match.

The uncomfortable truth is that most SEA SMEs are one phishing email away from a crippling breach. Awareness campaigns and government advisories are not enough. What the region needs now is affordable, accessible, and enforceable cyber hygiene, before the next wave of attacks makes that conversation moot.

Read the full article.

REGIONAL

SEA raises US$3.07B via IPOs in H1 2026: Southeast Asia’s IPO market posted strong first-half numbers, with Deloitte data showing a broad recovery across the region’s exchanges driven by consumer, technology, and industrial listings.

Singapore’s manufacturing surges 12.2% on AI demand: AI infrastructure buildout is fuelling Singapore’s broader economic expansion, with Q2 2026 GDP growing 5.7% as electronics and precision engineering output climbed sharply.

PolicyStreet lifts Series C to US$26M to target gig workers: The Malaysian insurtech extended its Series C round to cover underserved gig economy workers and SMEs across Southeast Asia, signalling growing investor appetite for inclusive financial protection products.

Alibaba, Mirae Asset back PixVerse’s US$439M Series C: The AI video generation startup secured backing from prominent Asian investors as it expands into AI-powered gaming content, reflecting deepening crossover between generative AI and entertainment.

SimpleAI secures US$10M debt facility for APAC acquisitions: The startup is using debt financing to roll up accounting firms across Asia Pacific, an unusual capital structure that bets on professional services consolidation as an AI adoption wedge.

Gobi Partners taps NTT to bridge Japan-SEA startup ties: The VC firm is leveraging NTT’s corporate network to channel Japanese enterprise deals and distribution partnerships toward its Southeast Asian portfolio companies.

e27 expands AI matchmaking via Sony Acceleration Platform: The new collaboration connects e27‘s startup network with Sony’s acceleration programme, using AI-driven tools to surface relevant partnership and investment opportunities.

Thales, Singtel launch global eSIM network for enterprises: The joint eSIM solution targets multinational businesses operating across borders, simplifying connectivity management and reducing reliance on physical SIM infrastructure for enterprise mobility.

---

INTERVIEWS & FEATURES

Korea’s ecosystem trains founders, not just funds them: Seoul’s startup support infrastructure has shifted toward capability-building programmes that prioritise founder education, mentorship, and market access over pure capital deployment, a model SEA ecosystems are watching closely.

US$3.7B was pledged to SEA climate tech: where did it go: A forensic look at the gap between climate finance commitments made to Southeast Asia and capital actually deployed, revealing structural barriers in bankability, policy risk, and local capacity.

Singapore and Taiwan eye a new window of opportunity: Tightening US-China trade tensions have created fresh grounds for Singapore-Taiwan collaboration in semiconductors, advanced manufacturing, and digital trade, but political caution may yet blunt the opportunity.

Global exposure alone won’t make Asian startups ready to scale: Founders who have studied or worked abroad still face a distinct set of proof points when expanding internationally, and many underestimate the operational gaps between ambition and execution.

---

INTERNATIONAL

Nadella warns companies over surface-level AI adoption: Microsoft’s CEO issued a stark caution to enterprises deploying AI without redesigning underlying workflows, arguing that layering AI onto broken processes will compound inefficiency rather than resolve it.

Anthropic localises Claude pricing for India: Anthropic has adjusted Claude’s pricing specifically for the Indian market — its second-largest globally — marking a strategic push to deepen penetration in price-sensitive, high-growth emerging markets.

OpenAI pushes ChatGPT deeper into family households: A new family plan signals OpenAI’s intent to expand beyond professional and enterprise users, betting that domestic AI adoption will drive the next wave of subscriber growth.

Payoneer opens India tech hub, ramps up hiring: The US fintech firm is scaling its India engineering and product team through a dedicated technology centre, reinforcing the subcontinent’s role as a critical offshore development base for global fintech operators.

---

CYBERSECURITY

Apple sues former employee over OpenAI data theft: A former Apple engineer allegedly exploited a rare software bug to exfiltrate confidential files after accepting a role at OpenAI, exposing the insider threat risks that accompany talent movement between AI rivals.

The wildest allegations in Apple’s trade secrets lawsuit: Apple’s legal filing against OpenAI goes beyond data theft, alleging systematic efforts to extract proprietary AI research, with implications for how tech firms manage security around departing staff.

SoftBank, OpenAI launch Japan cybersecurity service for 3,000 firms: The joint offering targets Japanese enterprises facing mounting AI-era threats, combining OpenAI’s models with SoftBank’s enterprise distribution to deliver AI-driven threat detection at scale.

When AI leaves the screen, security becomes a product problem: As AI moves into physical and embedded systems, responsibility for cybersecurity shifts from IT departments to product teams — a structural change most companies are unprepared for.

Thailand’s scam epidemic is fundamentally a technology failure: Online fraud in Thailand has reached epidemic scale not because of a lack of laws, but because platforms, telcos, and regulators have failed to deploy available technical countermeasures effectively.

---

SEMICONDUCTOR

Nanya plans US$6B spend in 2027 riding AI memory boom: Taiwan’s Nanya Technology is committing substantial capex to expand DRAM capacity as AI workloads drive sustained demand for high-bandwidth memory across data centre deployments.

Intel kicks off US$5.7B chip plant expansion in Ireland: The US chipmaker has broken ground on a major fab expansion in Ireland, reinforcing its European manufacturing footprint as Western governments push to reduce reliance on Asian chip supply chains.

Bosch begins semiconductor production at first US plant: The German industrial giant has started sample production at its inaugural American chip facility, targeting automotive and industrial applications as onshoring momentum in US semiconductor manufacturing accelerates.

South Korea flags record 2027 budget as chip revenues surge: Seoul is planning its largest-ever budget at over US$530B, buoyed by booming semiconductor export revenues — with major allocations earmarked for AI infrastructure and chip R&D competitiveness.

Why APAC’s next infrastructure boom hinges on power diversification: Data centres and AI chip clusters across Asia Pacific are outpacing grid capacity, making energy diversification across nuclear, solar, and storage the defining constraint on the region’s digital infrastructure ambitions.

---

AI

Agnes AI launches 2.5 Flash, teases Pro model and Code app: The Singapore-based AI lab released a faster, lighter model variant while signalling an upcoming flagship and a standalone coding application, expanding its product surface across developer and enterprise use cases.

Most AI pilots die in week six. LinqAlpha does it differently: The enterprise AI firm argues that pilot failures stem from misaligned success metrics and change management gaps, not technical shortcomings, and has built its deployment methodology around that diagnosis.

---

THOUGHT LEADERSHIP

US crypto regulation: the 400-page rule that could define the industry: A sweeping new US regulatory framework for digital assets has arrived, and its implications stretch beyond American borders, potentially setting the global compliance baseline that SEA crypto firms will be benchmarked against.

Crossing the valley of death in deep tech commercialisation: Many promising deep tech startups collapse not at ideation or early traction, but in the commercialisation gap between proof of concept and revenue-generating scale, a challenge Southeast Asia’s ecosystem is only beginning to address systematically.

How to build an internal AI academy that actually works: Organisations deploying AI at scale need structured internal capability-building programmes; this strategic guide outlines the design principles, governance models, and learning architectures that separate effective AI academies from box-ticking exercises.

Most GTM failures are architecture problems in disguise: Startups misdiagnose go-to-market failures as strategic missteps when the root cause is usually structural, poorly sequenced channels, misaligned incentives, and organisational design flaws that no pivot can fix.

Stop calling automated broken processes AI transformation: Slapping AI onto dysfunctional workflows does not constitute transformation; it accelerates existing failures. True AI transformation requires process redesign before automation, not instead of it.

Seasonal product cycles: why timing shapes feature success: Product teams often overlook the temporal dimension of feature launches — this analysis argues that release timing relative to user behaviour cycles can determine adoption outcomes as much as product quality itself.

Bitcoin holds above US$65,000 after CPI release: Crypto markets digested a key US inflation print with relative calm, as Bitcoin maintained its level and analysts assessed whether macro tailwinds remain strong enough to sustain the current rally.

The post Ecosystem Roundup: Ransomware’s easiest targets are hiding in plain sight across SEA appeared first on e27.

Ransomware attacks against small- and medium-sized enterprises (SMEs) in Southeast Asia rose in the first quarter of 2026, according to new data from Kaspersky, underscoring how smaller companies remain exposed even as cybercriminal groups sharpen their tactics and shift towards more layered extortion models.

The cybersecurity company said 3.51 per cent of SMEs in Southeast Asia within its ecosystem were targeted by ransomware in Q1 2026, up from 2.92 per cent in the same period last year. While the increase is not dramatic, it points to a persistent problem for a segment of the economy that often lacks the budget, staff and technical maturity of large enterprises.

Also Read: Why cyber resilience is the new standard for SME survival

In Singapore, the proportion of SMEs targeted rose to 0.69 per cent from 0.57 per cent a year earlier. Malaysia saw a larger increase, from 2.09 per cent to 2.74 per cent, while Indonesia rose from 2.83 per cent to 4.01 per cent.

Kaspersky’s dataset also included India, where the figure climbed from 3.18 per cent to 4.07 per cent.

The Philippines, Thailand, and Vietnam recorded declines. The Philippines fell from 2.46 per cent to 1.80 per cent, Thailand from 1.28 per cent to 1.12 per cent, and Vietnam from 2.91 per cent to 2.56 per cent. But the broader pattern remains one of steady exposure rather than a retreat in attacker interest.

Why SMEs remain exposed

The numbers matter because SMEs form the operating backbone of Southeast Asia’s economy. Across ASEAN, micro, small and medium-sized enterprises account for the vast majority of businesses and a substantial share of employment. That scale makes them attractive targets for cybercriminals: individually less defended than large corporates, but collectively deeply embedded in supply chains, payments networks, and customer data flows.

Ransomware has also changed. The older model of encrypting files and demanding payment has been supplemented by “double extortion”, where attackers steal data before locking systems and then threaten to publish it if victims refuse to pay. That shift increases the pressure on smaller companies, which may be less able to absorb operational downtime, reputational damage or regulatory scrutiny.

Kaspersky noted that its detection metric captures only part of the attack chain. A ransomware incident typically involves several stages, including initial access, reconnaissance, privilege escalation, lateral movement and data exfiltration. The final deployment of the encryption Trojan is only one part of that process. If an attack is stopped earlier, it may not appear as a crypto-ransomware detection.

That means the actual level of ransomware activity around SMEs is likely higher than the headline figures suggest.

Fedor Sinitsyn, a security expert at Kaspersky, said backups alone are no longer enough. “Most modern ransomware actors use the double extortion approach, where the attackers not only encrypt the victim’s files, but also exfiltrate confidential data and threaten to leak it in case of non-payment,” he said. “A layered cyber protection strategy is hence needed to provide adequate protection from attacks.”

Leak sites reveal a crowded attacker market

Kaspersky’s Q1 2026 malware report also tracked ransomware groups by the number of victims added to dedicated leak sites. Clop topped the list, accounting for 14.42 per cent of victims published on the sites monitored by Kaspersky. Qilin followed with 12.34 per cent.

Also Read: 10 reasons not to pay the ransom in a ransomware attack

A newer group, The Gentlemen, ranked third. Kaspersky said the group emerged around July 2025 and has since expanded rapidly. Its tactics reportedly include custom-built tools for covert information gathering inside victim systems before ransomware deployment. The group is also believed to work with initial access brokers, who sell access to compromised organisations.

That model has become central to the ransomware economy. Rather than breaking into every target themselves, ransomware operators can buy access, rent infrastructure, outsource negotiations and use leak sites to pressure victims. This specialisation lowers the barrier to entry and makes it harder for defenders to treat ransomware as a single type of malware problem.

The competitive landscape in cybersecurity has become correspondingly crowded. Kaspersky competes with global vendors such as CrowdStrike, Palo Alto Networks, Microsoft, SentinelOne, Sophos, Fortinet, Trend Micro and Check Point, many of which are pushing endpoint detection and response, managed detection and response, and extended detection platforms to mid-market customers.

For Southeast Asian SMEss, however, the issue is often less about product availability than implementation. Many firms struggle with patch management, identity controls, endpoint visibility, employee training and incident response planning. Even when tools are deployed, they may not be monitored continuously.

The regional stakes are rising

The ransomware problem intersects with broader digitisation across Southeast Asia. As companies adopt cloud software, digital payments, e-commerce channels and remote work systems, their attack surface expands. Regulators are also tightening expectations around data protection and cyber resilience.

Singapore has taken a more structured approach through the Cyber Security Agency of Singapore and sector-specific requirements. Malaysia, Indonesia, Thailand, Vietnam and the Philippines have also been strengthening cybersecurity and data protection frameworks, though enforcement and organisational readiness vary widely.

Industry data suggests ransomware remains a material global threat. Verizon’s 2024 Data Breach Investigations Report found ransomware was present in 32 per cent of all breaches it analysed. IBM’s 2024 Cost of a Data Breach report put the global average cost of a breach at US$4.88 million, a level that would be existential for many smaller companies even if regional costs vary.

Those figures help explain why attackers continue to pursue SMEs. A small manufacturer, logistics provider, clinic, accounting firm, or software vendor may not be the ultimate prize, but it can offer a route into larger customers or critical supplier relationships. In Southeast Asia’s highly interconnected business environment, that makes SME cybersecurity a wider economic concern, not merely an internal IT issue.

Adrian Hia, Managing Director for Asia Pacific at Kaspersky, said attackers increasingly see SMEs as an entry point into broader supply chains. “They are also being directed at firms that often lack the resources to maintain dedicated cybersecurity teams or implement comprehensive patch management programmes, making them attractive targets for threat actors.”

Kaspersky’s recommendations are familiar but still unevenly adopted: keep software updated, monitor lateral movement and outbound traffic, maintain offline backups, deploy endpoint detection tools, and develop an incident response plan that includes supplier compromise.

Also Read: Thailand is suddenly on the frontline of a new ransomware wave

For SMEs, the challenge is prioritisation. Few can replicate enterprise-grade security operations. But the basics — patching critical systems, enforcing multi-factor authentication, testing backups, limiting privileges and knowing who to call during an incident — can meaningfully reduce risk.

The latest figures do not suggest a sudden ransomware crisis in Southeast Asia. They point to something more durable: a persistent, professionalised threat market that continues to find smaller businesses profitable. For a region whose digital economy depends heavily on SMEs adoption, that is a risk founders, operators and investors can no longer treat as a back-office concern.

 

The post Southeast Asian SMEs remain soft targets as ransomware groups refine extortion tactics appeared first on e27.

LinqAlpha, the New York-headquartered AI startup building intelligence tools for institutional investors, recently raised US$22 million in a Series A round anchored by AVP, Atinum Investment, and GFT Ventures, with a notably Asia-heavy syndicate including SV Investment, Mirae Asset, Samsung Securities, East Ventures, and others spanning Singapore, Hong Kong, South Korea, Japan, and India.

Founded by Jacob Choi, Subeen Pang, Jin Kim, and Hojun Choi — a team of former Goldman Sachs analysts and MIT computer science PhDs — LinqAlpha says its multi-agent platform is already used by more than 70 financial institutions across the US, Europe, and Asia, including buy-side clients such as Causeway Capital Management and Schonfeld Strategic Advisors, collectively managing over US$5 trillion in assets.

The company positions itself against both entrenched incumbents like Bloomberg and LSEG, and a crowded field of AI challengers such as AlphaSense, Hebbia, and Rogo, betting that persistent, firm-specific reasoning — not just faster search — is where the real edge lies.

Also Read: LinqAlpha raises US$22M to bring agentic AI to public-market investors

We spoke with Jin Kim, co-founder and Head of Forward Deployed Engineering about the fundraise, the company’s Asia strategy, and how LinqAlpha plans to compete.

Edited excerpts:

Your US$22M raise features an Asia-heavy syndicate — SBI, Mirae Asset, Samsung Securities, East Ventures. Deliberate strategy or following traction?

Both. Roughly half our revenue comes from Asia Pacific, so the capital base mirrors the client base. But the syndicate was deliberately built: in institutional finance, investors are also distribution. Firms like SBI, Mirae Asset, and Samsung Securities are operating institutions in the markets we serve; that alignment shortens trust-building cycles that normally take years. We didn’t raise Asian capital to enter Asia later; global coverage, Asian languages, and multi-asset support were in the design from Day One.

You’re headquartered in New York, yet clients and capital skew Asia. When does it make sense to shift your center of gravity to Singapore or Tokyo?

We, at LinqAlpha, run a distributed model rather than one centre: New York for business development, Seoul as our product/engineering hub, subsidiaries in Hong Kong and Singapore, with London next. Singapore is where we’ve made the on-the-ground commitment, a dedicated local team, not a sales outpost, but people who co-design deployments with regional institutions. The timing isn’t accidental: MAS just launched the Future of Finance Institute to move AI in financial services from experimentation to deployment. That deployment gap is our entire business.

You count 70+ financial institutions as clients, but client count can be a vanity metric. How embedded is LinqAlpha in daily workflows, and how do you measure it?

We agree it’s vanity, which is why we manage for depth across three layers:

1. Daily-workflow usage: morning briefings, alerts, meeting-prep agents that fire before the user’s day starts, not ad-hoc Q&A
2. Expansion within accounts: trials converting to multi-seat, multi-team deployments
3. Integration depth: clients moving from app to API access, building our agents into their own systems

The pattern we watch: users going from reading our output to relying on it—starting with a briefing, pushing results to PMs, then asking us to build custom trackers for signals nobody else covers, often in Asian-language sources their other tools can’t read.

Hebbia, AlphaSense, Rogo, Dataminr — plus Bloomberg and LSEG embedding AI into terminals. Why should a CIO choose LinqAlpha over waiting for incumbents to catch up?

We’re solving different problems. Bloomberg and LSEG are indispensable data infrastructure; AlphaSense built a strong content library with AI on top. But a platform selling the same content to everyone will, by design, give every subscriber the same AI answer from the same corpus.

What a CIO competes on is the firm’s own frameworks, thesis history, and internal research. We encode that — per firm, permissioned, isolated — in what we call a second brain. Ask two funds “what are the top AI trades today” and a generic tool names the same mega-caps for both. Our platform answers in the context of each firm’s own mandate. We sit on top of data clients already license and reason across asset classes — equities, macro, credit, FX –in one connected system rather than siloed products.

AI hallucinations can directly influence capital allocation. What safeguards ensure accuracy and auditability, and has a failure ever cost a client?

We designed the platform assuming models are fallible, so safeguards are architectural. Every claim is grounded in licensed, vetted data with a citation back to source, auditable in one click. Numbers are computed deterministically through code, not generated by an LLM. We run multiple frontier models and neutralise their individual biases; our research on systematic sector/style biases in base models was accepted at ICLR and presented at a BlackRock quant conference.

Also Read: In the age of AI, people matter more than ever

We call this discipline harness engineering: as models get more powerful, value shifts to controlling them — permissioning, audit trails, human-in-the-loop checkpoints built for regulated finance. On failures: we have not had an incident where a platform error drove a capital loss for a client.

Clients feed you proprietary research and conviction signals. How do you handle data security?

Isolation isn’t a feature; it’s the product’s precondition. Each firm’s knowledge layer is siloed: notes, theses, and research never train shared models and never cross accounts. One client’s second brain is architecturally invisible to every other client’s.

For institutions wanting an even tighter boundary, our API and MCP deployment patterns let agents operate inside the client’s own environment, so sensitive context need not leave their perimeter. The incentive structure matters too: our business model is per-firm value, not data aggregation. Security reviews and regulatory addenda with global banks are table stakes, and we treat them as part of the product.

Southeast Asia is fragmented — multilingual, multi-regulatory, inconsistent data. How does the platform handle Bahasa Indonesia filings, Thai regulatory announcements, and Vietnamese commodity flows simultaneously?

That fragmentation is the inefficiency we were founded to arbitrage. The platform analyzes 20,000+ companies across 80+ markets in 20 languages, reading local-language primary sources natively rather than waiting for English translations that arrive late or never.

In Asia, that’s where alpha lives: information that’s public but not yet priced because it sits behind a language barrier. Clients already track signals in Chinese-, Korean-, and Japanese-language sources that English-first platforms structurally miss; the same architecture extends across Southeast Asia. Equally important is multi-asset design: an Indonesian commodity signal reads through to Singapore-listed equities, regional FX, and credit in one connected graph. Where coverage needs deepening, we build it hand-in-hand with regional clients—that’s the global best practice we’re bringing to Singapore, not a US product with a Singapore price list.

With a Berkeley MFE/Goldman/MIT pedigree, doors open easily, but institutional adoption is slow. What’s been the biggest obstacle converting pilots to long-term contracts?

Never model quality in a demo. The real obstacle is earning a place in daily workflow at a conservative institution—what kills pilots industry-wide is a tool that impresses in week one and is forgotten by week six. We treat every trial as an implementation, instrumenting adoption user by user and workflow by workflow.

The second obstacle is institutional trust: security review, compliance sign-off, data governance. We stopped treating that as friction and started treating it as the sale, because the risk owner is usually the real buyer. What converts pilots is co-designing an AI roadmap with client leadership over the next one to two years, rather than selling seats.

Buy-side clients manage US$5 trillion+ in assets, striking for a Series A. What’s your pricing model? Recurring SaaS, or a services-heavy business that doesn’t scale?

It’s recurring software by design: seat- and entitlement-based subscription with enterprise tiers for API access and advanced modules. The same platform serves a hedge fund pod and a bank’s research floor. The insight most people miss: the “bespoke” part—learning each firm’s framework—is performed by the system itself. The second brain is built by agents from the client’s own permissioned data, not consultants billing hours. That makes personalization compound instead of costing more. The AUM figure is a statement about who trusts us, not a revenue multiplier—but reference clients at that tier are the moat, because institutional buyers follow institutional proof.

If AI “changes what analysts can know,” doesn’t wide adoption commoditise the very edge you promise?

That critique is fatal for generic AI. This is exactly why we built the opposite. If every investor used the same model on the same data, the edge would be arbitraged away in a quarter. Our architecture inverts it: agents reason in the context of each firm’s own thesis history and mandate, so two funds asking the identical question get different, both correct, answers—the platform amplifies different brains. Adoption doesn’t converge outputs; it compounds each firm’s accumulated judgment.

Also Read: Is generative AI the game-changer for productivity?

Think of Bloomberg in the 1990s: everyone had the terminal, yet returns diverged wildly, because the edge was never the tool; it was what each firm did with it. We’ve made “what each firm does with it” the product itself. AI is shifting the scarce resource from information access to quality of questions and speed of connecting dots. The real risk for a CIO isn’t adopting AI too early; it’s letting a competitor’s second brain start compounding a year before yours does.

The post Most AI pilots die in week six. Here’s what LinqAlpha does differently appeared first on e27.

The rapid adoption and expansion of artificial intelligence is reshaping global infrastructure demand. From hyperscale data centres to advanced semiconductor fabrication facilities, AI-led growth is driving an unprecedented need for mission-critical assets across Asia Pacific (APAC).

APAC’s power infrastructure is struggling to keep pace. These facilities are highly power-intensive, requiring uninterrupted, high-reliability energy to maintain operations and meet stringent uptime requirements. Mature markets in APAC face grid saturation and limited space for new generation, while emerging markets often contend with reliability issues, transmission bottlenecks or regulatory complexity. The scale and speed of digital and industrial development now risk outstripping expansion in generation, transmission, and distribution capacity.

The next phase of infrastructure growth in APAC will be defined not just by demand, but by how effectively developers plan for and secure power. In an environment of tightening supply, rising costs and increasing volatility, diversification and early-stage power strategy are becoming critical determinants of project success.

The power squeeze

Recent geopolitical developments, such as the conflict in the Middle East, have further amplified these constraints. Linesight’s analysis shows that the impact on APAC is particularly acute. The Strait of Hormuz remains one of the world’s most critical energy chokepoints. Around 25 per cent of global seaborne oil trade and about 20 per cent of global LNG trade move through the Strait, with around 80 per cent of oil flows destined for Asia. Many APAC economies rely on imported oil and LNG, increasing their exposure to supply shocks and cost escalation.

At the same time, the power demands of mission-critical facilities continue to climb. Across markets such as Singapore, Malaysia and Japan, data centre pipelines are expanding rapidly, while governments are actively onshoring semiconductor capacity to strengthen supply chain resilience. Singapore’s Green Data Centre Roadmap, for example, aims to provide at least 200MW of additional near-term capacity (in its second data centre call for applications), while Malaysia’s data centre load could exceed 5,000MW by 2035. This convergence is placing unprecedented pressure on already constrained power systems.

In several jurisdictions, power connection timelines now exceed typical construction programmes, creating a material risk to project viability. With competition for limited power capacity intensifying against a backdrop of continued geopolitical uncertainty, the construction sector is confronting an unforgiving reality: developments that fail to secure resilient and diversified power strategies early risk delay, de-scoping or obsolescence. Power availability should be treated as an early-stage project risk, not a late-stage utility consideration.

Also Read: The AI-quantum collision: Navigating the 2026 infrastructure inflection point

The case for diversified power strategies

In this environment, power diversification is no longer just a strategic consideration; it is becoming as important as land, labour and capital. Reliance on a single fuel source or grid connection exposes projects to volatility, while diversified power strategies provide optionality and stability when systems are under stress.

Across APAC, developers are increasingly integrating alternative energy sources to build a more resilient energy mix and buffer against supply constraints or price shocks. On-site solar generation, battery energy storage systems, microgrids, waste-to-energy solutions and long-term power purchase agreements are being deployed to stabilise supply and manage operating costs. These are generally hybrid solutions, where the combination of grid power, renewables and backup generation enables developments to proceed where grid capacity alone would have been insufficient, whilst maintaining uptime during periods of heightened disruption.

Lower-carbon power as a key part of the diversification playbook

Beyond diversification itself, a critical consideration is the type of power integrated into these strategies, especially given APAC’s decarbonisation ambitions. Governments have announced capacity targets supported by policy reform and investment incentives. Singapore is targeting at least 2GWp of solar deployment by 2030 and, by 2030, expects at least nine hydrogen-compatible gas-fired power plants as part of its longer-term low-carbon transition. Japan’s latest energy policy points to renewables reaching 40 to 50 per cent of its power generation mix by FY2040, supported by targets including 30 to 45GW of offshore wind by 2040.

Diversified and lower-carbon power strategies are no longer just environmentally desirable; they are commercially compelling. Beyond capacity, rising costs associated with carbon-intensive energy are accelerating the case for transition. Carbon pricing mechanisms and reduced subsidies are pushing up long-term costs, while incentive schemes and concessional financing are steadily improving the economics of renewables.

For developers, this shift opens a broader alternative playbook. Distributed energy resources, such as rooftop solar and behind-the-meter storage, can reduce grid dependence and accelerate time to power. Corporate power purchase agreements can provide long-term price certainty while supporting new renewable projects. In select markets, private microgrids and dedicated generation assets are emerging as viable solutions for energy-intensive developments. Diversification into lower-carbon energy not only supports customer commitments around uptime, emissions reduction and supply chain transparency, but also strengthens their ability to attract increasingly environmentally minded occupiers and investors, while reducing exposure to future ESG-related regulatory costs.

Also Read: The rise of AI twins: From assistant to infrastructure

However, these alternatives require early integration into project planning. Site selection, land availability, grid adjacency and permitting regimes all influence the feasibility of diversified power solutions. Successful delivery increasingly depends on aligning technical, commercial and regulatory considerations from the earliest project phases. Optimal solutions will also vary by market. Mature, land-constrained locations may prioritise imported renewable energy, off-site PPAs and energy efficiency, while emerging industrial hubs may have more scope for on-site or near-site generation and private power infrastructure.

What will define winners in APAC’s next infrastructure boom

Power diversification is no longer an environmental or ethical choice alone; it is an operational necessity for mission-critical infrastructure across APAC. In a region defined by rapid growth, constrained grids and geopolitical uncertainty, diversified power strategies provide stability, resilience and confidence.

Assets that integrate multiple power sources benefit from more predictable uptime, stronger investor appeal and enhanced long-term value. To fully realise these advantages, diversification must be embedded from the outset, influencing site selection, design development and capital strategy.

The winners in APAC’s next phase of infrastructure growth will be those that recognise power as a strategic asset, planned early, costed accurately, managed actively and diversified intelligently.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.

The post Why the next infrastructure boom in APAC depends on power diversification appeared first on e27.

Product teams are trained to ask familiar questions. Who is the user? What problem are we solving? How often does it occur? How painful is it? What is the willingness to pay? Those questions matter, but there is another one that quietly shapes adoption far more than many teams admit.

When does this problem actually become real in the customer’s world?

That is a different question from frequency. It is not asking whether a problem exists in principle. It is asking when the problem becomes urgent enough, visible enough, or costly enough for a feature to earn attention, budget, workflow change, and repeat use.

Many features fail not because they are weak, but because they are mistimed. They arrive outside the window where the customer is ready to care. Then the team misreads the result. It concludes the feature lacked value when the deeper issue was that the value did not line up with the customer’s calendar.

Most feature adoption is seasonal in ways teams do not name

When people hear seasonality, they usually think of obvious consumer patterns. Retail peaks in holidays. Travel surges in summer. Fitness spikes in January. Those are real, but they are only the visible end of the idea.

In practice, many products live inside less obvious seasons.

Enterprise products have planning seasons, budgeting seasons, procurement seasons, audit seasons, renewal seasons, hiring seasons, transformation seasons, and risk seasons. Internal tools live through quarter-end pressure, annual planning rituals, compliance reviews, and leadership changes. Even collaboration features can behave seasonally because teams communicate differently during launches, restructures, onboarding waves, or periods of cost control.

The feature itself may not change. The customer’s readiness to adopt it does.

Also Read: Product DNA testing: How features inherit traits from parent products

Some features are not evergreen, and that is fine

One of the unhelpful biases in product thinking is the assumption that the best features behave like evergreen assets. They should show steady demand, broad applicability, and consistent usage. That expectation sounds rational, but it can distort judgement.

Some features are not meant to be used evenly. Their value comes from intensity, not constancy.

A planning tool may matter enormously during one month and sit nearly dormant during others. A compliance workflow may become critical during review periods and almost invisible in quieter quarters. A feature that supports hiring, onboarding, migration, or renewal may deliver huge value in concentrated windows rather than through daily engagement.

That does not make the feature weak. It makes it cyclical.

The real mistake is evaluating cyclical value through non-cyclical expectations. Teams look at monthly usage and panic because the graph is uneven. They ask whether the feature is sticky enough, when the better question is whether it becomes indispensable at the exact moment it should.

This is where product maturity shows. 

The market has calendars, even when your roadmap ignores them

Most roadmaps are built around internal logic. Engineering capacity, strategic themes, executive priorities, dependencies, and quarterly planning all shape what gets released when. That is understandable, but it often means the product launches according to the company’s calendar rather than the customer’s.

This is one of the least discussed reasons good features underperform.

A team may launch a budgeting capability in the quarter after customers set budgets. It may introduce governance controls after the compliance window has passed. It may ship a staffing feature after hiring freezes begin. It may release operational tooling during the busiest commercial period, when nobody has the attention to absorb process change, no matter how sensible the new workflow looks in a demo.

The product team then spends weeks trying to work out what went wrong in the positioning, onboarding, or interface. Sometimes the answer is far simpler. The feature reached the market at a time when the customer had no spare bandwidth, no urgent reason to switch, or no practical ability to act.

Also Read: The problem with ‘PM as CEO of the Product’: A myth that hurts more than helps

A badly timed launch can produce false negatives

This is one of the more expensive product mistakes because it leads to the wrong learning.

A feature launches. Adoption is weak. Leadership loses confidence. The team trims investment, shifts attention, or decides the market is not ready. In some cases, that judgement is right. In many others, it is premature.

The problem is that timing failures often masquerade as product failures.

If a capability is introduced outside the season in which customers are willing to act, the team collects weak signals. Low usage, slow setup, muted excitement, limited word of mouth. Those signals look like poor product-market fit, but they may actually reflect poor temporal fit.

This matters because the remedy is different. A weak product needs redesign. A mistimed product may need reintroduction, better sequencing, stronger preparation, or a different commercial motion around the same underlying capability.

The danger is that teams abandon the right idea after reading the wrong evidence.

Features have windows of activation, not just user segments

Most product strategy frameworks focus on segmentation by user type, company size, industry, geography, or maturity. Those still matter, but they are not enough. Features also need to be segmented by time.

A more serious product question is not only who this feature is for, but when it is most likely to activate.

That changes how you think about rollout, education, pricing, and success measurement. If a feature only matters in planning season, then awareness needs to exist before planning season, not during or after it. If the capability becomes crucial during audits, then setup and training need to happen in the quieter period before the audit window opens. If a workflow matters only at renewal, then the product cannot wait until the renewal moment to explain its value.

In other words, teams need to design for the activation window, not just the feature itself.

This is where many companies underinvest. They build the capability and assume timing will sort itself out. It rarely does. Time needs orchestration in the same way as functionality.

Also Read: The systemic minimum effective dose: Redesigning productivity through precision

The strongest features often prepare long before they are used

This is another subtle but important point. A feature’s moment of highest value is not always the same as its moment of highest preparation.

Take any capability that supports a critical but infrequent workflow. The actual use may happen in a compressed, high-stakes period, but the product work that enables successful adoption has to begin much earlier. Permissions have to be configured. Data needs to be clean. Users need to understand why the feature exists. Teams need to trust that it will hold up when the important moment arrives.

That means some of the most important product work sits before the season, not inside it.

Weak product teams focus only on the event. Stronger ones think about readiness. They understand that the feature is not being adopted in the same moment it is being used. It is being adopted in the months when the customer decides whether to rely on it later.

That distinction is hugely important for enterprise products, operational tools, financial workflows, and anything that carries risk. Customers do not place trust instantly on the day of urgency. They decide beforehand what they are willing to trust when urgency arrives.

—

Editor’s note: e27 aims to foster thought leadership by publishing views from the community. You can also share your perspective by submitting an article, video, podcast, or infographic.

The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of e27.

Join us on WhatsApp, Instagram, Facebook, X, and LinkedIn to stay connected.

The post Seasonal product cycles: Why some features only work at certain times appeared first on e27.